Downloaded from genome.cshlp.org on October 7, 2021 - Published by Cold Spring Harbor Laboratory Press An Efficient and Optimized PCR Method with High Fidelity for Site-directed Mutagenesis Qianwa Liang, Lisha Chen, and Armand J. Fulco Department of Biological Chemistry and the Laboratory of Structural Biology and Molecular Medicine, School of Medicine, University of California, Los Angeles, California 90024-1737 We have developed an efficient provements in efficiency, flexibility, round of PCR amplification, the muta- method for site-directed mutagene- and fidelity. genic primer and the antiparallel univer- sis using two subsequential rounds of sal primer are used to produce a specific PCR. In this method, PCR conditions DNA fragment carrying the mutation. are optimized to favor high fidelity of This DNA fragment is purified and then Taq DNA polymerase in the presence PCR with Taq DNA polymerase has used as a primer together with the sec- of equimolar concentrations of been widely used for both the amplifica- ond universal primer for the second MgCI 2 and dNTP in the reaction mix- tion of specific DNA sequences and site- round of PCR amplification. However, a ture (pH $.5--6.2). This method directed mutagenesis. In the last 6 years, frequent drawback of this method, as makes use of a pair of universal prim- two major methods for site-directed mu- pointed out by Barettino et al., ~4~ is inef- ers and the multiple cloning site of tagenesis using two subsequential ficient priming by the megaprimer frag- pUC/M13 vectors. Only one muta- rounds of PCR amplification have been ment in the second round of PCR ampli- genic primer is required per target developed. The first method, designated fication, which results in very low yields site. In the second round of PCR, the overlap extension, was described by Ho of the full-length fragment of target 3' extension of the wild-type DNA et al. (1~ In this method a pair of overlap- DNA. Recently, several different ap- strand is blocked by the presence of a ping mutagenic primers and two flank- proaches for improving the megaprimer segment of nonhomologous se- ing primers are required. With two dif- method have been described (4's~ but, as quence at its 3' end, and as a conse- ferent combinations of one flanking the investigators, note, these approaches quence, the amplified, full-length primer and one mutagenic primer, two introduce their own limitations. More- DNA fragment is chiefly from the mu- DNA fragments having overlapping ends over, in reviewing the various proce- tant strand. Furthermore, because are generated by two separate PCR am- dures for PCR-mediated site-directed the mutated DNA fragment has plifications. These fragments are com- mutagenesis published to date, we found flanking restriction sites different bined in a subsequent "fusion" reaction that little attention had been paid to the from those of the wild-type DNA frag- in which the overlapping ends anneal to optimization of mutagenic PCR with re- ment, the wild-type DNA fragment is each other, allowing the 3' overlap of gard to the fidelity of Taq DNA poly- totally excluded in the step involving each strand to serve as a primer for the 3' merase. Should DNA synthesis by Taq selective cloning of the mutant DNA extension of the complementary strand. DNA polymerase, used for this purpose, fragment. This method was success- The resulting fusion product is amplified be carried out under low-fidelity reac- fully used to introduce four, nonad- further by PCR. In practice, however, we tion conditions, the occurrence of un- jacent mutations in the S' regulatory found that the annealing of overlapping desired mutations would become a ma- region of the cytochrome P450SM_~ ends often fails to occur, which may be jor concern. gene. All 20 analyzed clones from attributable to the complementary reas- On the basis of the megaprimer strat- these four cases of mutagenesis car- sociation of double-stranded DNA in the egy, we developed an efficient method ried the desired mutations, and no fusion reaction and the adoption of sec- for site-directed mutagenesis using PCR undesired mutations were observed. ondary structure in single-stranded DNA for the introduction of multiple desired We observed that the larger the num- produced in the subsequent PCR. The mutations in the 5' regulatory region of ber of mismatched nucleotide resi- second method termed megaprimer, was the cytochrome P450BM_3 gene. There are dues in the mutagenic primer, the developed by Landt et al., (2~ based on the three major advantages in the method higher the concentration of MgCI 2 scheme described by Kammann et al. ~3) that we present here. First, the extension was necessary for successful PCR am- This method requires just one muta- of the 3' end of wild-type DNA strand is plification. Our experimental results genic primer and two flanking universal blocked by the introduction of a seg- indicate that this method offers im- primers for pUCfM13 vectors. In the first ment of nonhomologous DNA sequence 4:269-2749 by Cold Spring Harbor Laboratory Press ISSN 1054-9803/95 $5.00 PCR Methods and Applications 269 Downloaded from genome.cshlp.org on October 7, 2021 - Published by Cold Spring Harbor Laboratory Press to its 3' end so that the full-length frag- into plasmid vector pTZ19R to yield a re- RESULTS ment produced by the second round of combinant plasmid designated pFL3-1. Optimization of Reaction Conditions PCR amplification is mainly from the 3' To introduce site-directed mutations in extension of the mutant strand. Second, the regulatory region, the first round of for Mutagenic PCR the mutated DNA fragment can be selec- PCR was set up in a total volume of 100 Minimizing the frequency of undesired tively cloned into a plasmid vector by I~l containing 1 Izl of pFL3-1 (-5 ng), 20 mutations in PCR-mediated site-directed using appropriate flanking restriction pmoles of each primer, 10 ixl of 10x re- mutagenesis can be achieved by maxi- sites, as the mutated DNA fragment has action buffer [500 mM KCI, 100 mM Tris- mizing the fidelity of the DNA poly- flanking restriction sites different from HCl (pH 9.0), 1.0% Triton X-100], and merase. The error rate of Taq DNA poly- those of the wild-type DNA fragment. Fi- 2.5 units of Taq DNA polymerase; opti- merase used in the PCR can be nally, the conditions of PCR for site-di- mized amount of equimolar concentra- influenced by many factors in the reac- rected mutagenesis have been optimized tions of MgCl z and dNTP (see Optimiza- tion; (1~ major factors include MgC12 with regard to fidelity of Taq DNA poly- tion of Reaction Conditions for level, dNTP concentration, and pH con- merase to minimize undesired muta- Mutagenic PCR, below) were added. The ditions. Eckert and Kunkel(11)evaluated tions. reaction mixtures were overlayed with these factors extensively in PCR and re- 50 Ixl of light mineral oil, and the reac- ported that high fidelity of Taq DNA MATERIALS AND METHODS tion was carried out in an automatic polymerase could be reached when thermal cycler for four cycles of 2 min at MgC12 and dNTP were present at equi- Materials 94~ 1 min at 48~ and 2 min at 72~ molar concentration or at a pH between Restriction endonucleases, T4 DNA li- and then for 20 cycles of 1 min at 92~ 5 and 6 (70~ at 1.5 mM MgC1 z. Because gase, and T4 DNA polymerase were pur- 1 min at 58~ and 2 min at 72~ The the PCR mixtures, when set up with the chased from either New England Biolabs 72~ incubation of the last cycle was ex- 10x reaction buffer supplied with Taq or GIBCO BRL. Taq DNA polymerase was tended for an extra 5 min before the re- DNA polymerase from Promega, gave a purchased from Promega. Oligonucle- action samples were cooled to room pH of 5.5-6.2 at 70~ our efforts to op- otide primers used in PCR and DNA se- temperature. The second round of PCR timize the mutagenic PCR conditions quencing were synthesized by Integrated amplification was set up as described with regard to the fidelity of Taq DNA DNA Technologies, Inc. Deoxynucle- previously, (7) except 2 mM MgClz and 0.5 polymerase were focused on the adjust- otide triphosphates (dNTPs) and mM each of the four dNTPs were used ment of the MgClz and dNTP concentra- [~-3SS]dATP were from Amersham Corp., here. The PCR was performed for 20 cy- tions. Mutagenic primers and pUC/M13 the Geneclean kit was purchased from cles, each cycle consisting of 92~ for 1 universal primers used for the introduc- Bio 101, Inc., and the Sequenase kit and min, 57~ for 1 min, and 72~ for 2.5 tion of mutations in three different loca- plasmid pTZ19R were from U.S. Bio- min. The last cycle was followed by 72~ tions of the 5' regulatory region of the chemical Corp. The 5' regulatory region for 5 min before cooling to room tem- P450BM_~ gene are shown in Table 1. The of the cytochrome P450BM_3 gene in Ba- perature. mutagenic primers carry one to nine cillus megaterium ATCC 14681 was mutant nucleotide residues. Because the cloned in our laboratory. (6) The DNA se- error rate of DNA synthesis by Taq DNA polymerase in the PCR has been shown quence of the P450BMo3 gene, including General DNA Methods the 5' regulatory region, is available un- to increase with increasing MgCl 2 con- der GenBank accession number J04832. DNA cloning, plasmid isolation, restric- centration (dNTP at 1 mM), O1) the min- Escherichia coli strain JM109 was used as tion enzyme digestion, and gel electro- imum concentration of MgClz required a host for plasmid transformation and phoresis were performed according to for efficient mutagenic PCR amplifica- preparation.